Effect of folate status and age on DNA methylation and gene expression

Abstract

DNA methylation patterns are malleable and are modifiable by environmental factors, including diet. Folate (a water-soluble B vitamin) is a major methyl donor for methylation of DNA and other cellular macromolecules. Since patterns of DNA methylation are established early in embryonic and fetal life, I hypothesised that reduced maternal dietary folate intake would alter DNA methylation patterns and corresponding gene expression in the progeny. A second hypothesis was that maternal dietary folate intake would modify effects of later dietary exposures on DNA methylation, in particular effects of a high fat diet. Third, I hypothesised that loci where DNA methylation was affected by dietary folate in utero and / or by dietary fat would also show age-related differences in DNA methylation and gene expression. This project aimed to investigate the effects of feeding a folate-depleted (0.4mg/kg) versus a folate adequate (2.0mg/kg) diet during pregnancy and lactation, and high fat (HF, 20%) versus low fat diet (LF, 5%) from weaning until 6 months of age on gene promoter methylation in offspring of C57BL/6J mice. DNA methylation at ten loci: Igf2-DMR1, Esr1, p16, Slc39a4-CGI1 & -CGI2, Obfc2a-amp1, -amp2 & -amp3, and Ppm1k-amp1 & -amp2 was measured by Pyrosequencing in fetal gut tissue. In adult offspring, methylation and expression of this gene panel was quantified in the colon and proximal small intestine (prox SI). Finally, the impact of folate status and age on ESR1, OBFC2A and SLC39A4 methylation and expression was investigated using macroscopically normal human colon mucosal biopsies. In fetal mouse gut, there was no effect of maternal folate supply on methylation at any of the CpGs investigated in Esr1, p16, Igf2-DMR1, Slc39a4-CGI2 and Obfc2a. However, DNA methylation at Slc39a4-CGI1 and Ppm1k-amp1 were significantly lower in gut DNA from folate depleted mice (ANOVA, p<0.05). In adult offspring, methylation at Slc39a4-CGI1 was significantly lower in the prox SI from mice fed the HF diet (ANOVA, p<0.05, n=24 for LF, n=24 for HF diet). In contrast, methylation at p16, Obfc2a-amp1, Obfc2a-amp2, Ppm1k-amp1 and Ppm1k-amp2 was higher in the prox SI of iii adult offspring fed the HF diet (p<0.05). HF feeding led to significantly lower expression of p16, Obfc2a and Ppm1k in the prox SI (p<0.05). For adult mouse colon, methylation at Esr1 and Obfc2a-amp2 was increased in those fed the HF diet whereas for Slc39a4-CGI1 and Slc39a4-CGI2, methylation was lowered in the HF group (p<0.005). And there was no detectable change in gene expression of this gene panel in colon as a result of HF feeding. In human colorectal mucosal biopsies, methylation of all 8 CpGs assayed in ESR1 was positively correlated with age (p<0.05, n=34) but there was no apparent age-related change in ESR1 expression (p>0.05). In summary, maternal folate depletion reduced methylation at Slc39a4 and Ppm1k in the murine fetal gut but this effect disappeared when mice were given the same folate supply from weaning. HF feeding altered methylation of Esr1, Slc39a4 and Obfc2a in adult mouse colon and Slc39a4, p16, Obfc2a, and Ppm1k in the prox SI. The effect of maternal folate depletion and HF feeding of offspring on DNA methylation appeared to be CpG, locus and tissue specific in mice. Expression of p16, Obfc2a and Ppm1k was lowered in the prox SI of HF fed mice. In the human colorectum, ESR1 methylation increased with age but there was no detectable effect on expression.